PDL1 is required for peripheral transplantation tolerance and protection from chronic allograft rejection

Abstract

The PD-1:PDL pathway plays an important role in regulating alloimmune responses but its role in transplantation tolerance is unknown. We investigated the role of PD-1:PDL costimulatory pathway in peripheral and a well established model of central transplantation tolerance. Early as well as delayed blockade of PDL1 but not PDL2 abrogated tolerance induced by CTLA4Ig in a fully MHC-mismatched cardiac allograft model. Accelerated rejection was associated with a significant increase in the frequency of IFN-gamma-producing alloreactive T cells and expansion of effector CD8(+) T cells in the periphery, and a decline in the percentage of Foxp3(+) graft infiltrating cells. Similarly, studies using PDL1/L2-deficient recipients confirmed the results with Ab blockade. Interestingly, while PDL1-deficient donor allografts were accepted by wild-type recipients treated with CTLA4Ig, the grafts developed severe chronic rejection and vasculopathy when compared with wild-type grafts. Finally, in a model of central tolerance induced by mixed allogeneic chimerism, engraftment was not abrogated by PDL1/L2 blockade. These novel data demonstrate the critical role of PDL1 for induction and maintenance of peripheral transplantation tolerance by its ability to alter the balance between pathogenic and regulatory T cells. Expression of PDL1 in donor tissue is critical for prevention of in situ graft pathology and chronic rejection.

FIGURE 1

A, Early PDL1 but not PDL2 blockade abrogates tolerance induction by CTLA4Ig. CTLA4Ig treatment induced long-term graft survival in all recipients (MST > 100 days, n = 8; p < 0.0001vs MST 7 days; n = 18 for IgG isotype-treated controls). Administration of anti-PDL1 mAb starting on the day of transplantation resulted in graft rejection in the CTLA4Ig-treated recipients (MST = 31.5 days; n = 12; p < 0.0005). In contrast, administration of anti-PDL2 mAb did not result in any significant change in graft survival in the CTLA4Ig-treated recipients (MST > 100 days; n = 5; p = ns). B, Delayed PDL1 blockade abrogates tolerance induction by CTLA4Ig. Although long-term allograft survival was achieved by CTLA4Ig treatment, administration of the anti-PDL1-mAb beginning on day 60 posttransplantation also precipitated rejection in CTLA4Ig treated recipients (MST = 83, n = 6, p = 0.02).

FIGURE 2

Pathological analysis of cardiac tissue after CTLA4Ig treatment 2-3 wk after PDL1 blockade. Histological evaluation of grafts harvested from PDL1 mAb-treated recipients 2 wk after blockade showed scattered inflammatory cell infiltrates, vasculitis with inflammation and variable degrees of vascular luminal obliteration. In comparison, grafts from CTLA4Ig-treated animals were protected from acute and chronic rejection.

FIGURE 3

A-C, ELISPOT analysis of Th1 (IFN- γ (A) and Th2 (B and C) cytokine profile of splenocytes derived from transplant recipients treated with CTLA4Ig 2-3 wk after PDL1 blockade. In the CTLA4Ig treatment group, the frequency of alloreactive IFN-γ-producing splenocytes was significantly decreased as compared with the untreated control recipients (A). Anti-PDL1-mAb administration during the induction phase resulted in significant increase of the frequency of alloreactive IFN-γ-producing splenocytes (A). CTLA4Ig treatment produced a slight but nonsignificant decrease in the frequency of IL-4-producing cells as compared with controls, whereas administration of anti-PDL1-mAb significantly increased the IL-4-producing T cell frequency (B). Interestingly, CTLA4Ig treatment significantly increased in the frequency of IL-5-producing cells as compared with controls (C), and anti-PDL1 mAb further increased the IL-5-producing T cell frequency as compared with the CTLA4Ig-treatment group (C). Data are representative of three independent experiments and indicate the mean of triplicate results in each experiment. D and E, Frequency of alloreactive CD8 effector T cells in splenocytes derived from transplant recipients treated with CTLA4Ig 2-3 wk after PDL1 blockade. The frequency of effector CD8⁺ T cells (expressing a CD62L^lowCD44^high phenotype) was significantly decreased in the CTLA4Ig treatment group as compared with controls (3.6 ± 1.0% vs 23.7 ± 0.4%, p < 0.001). Anti-PDL1 mAb significantly increased the frequency of effector CD8⁺ T cells, as compared with the CTLA4Ig treatment group (12.1 ± 2.5% vs 3.6 ± 1.0%, p < 0.001). A representative experiment is shown (D).

FIGURE 4

Immunohistochemical analysis of FOXP3 in cardiac tissues after CTLA4Ig and PDL1 blockade. Foxp3 staining in the heart grafts demonstrated significant increase in the percentage of infiltrating cells expressing Foxp3 in CTLA4Ig-treated recipients. Importantly, PDL1 blockade resulted in significant decrease in the percentage of cells expressing Foxp3 as compared with CTLA4Ig treatment alone (25 ± 2.2% vs 11.6 ± 4.6, n = 6, p = 0.02).

FIGURE 5

Role of recipient versus donor PDL1 expression after treatment with CTLA4Ig. A, PDL1-deficient recipients demonstrated significantly shortened graft survival after CTLA4Ig therapy (MST = 30, n = 5, p = 0.0007) as compared with wild-type B6 recipients. B, PDL1-deficient heart allograft were accepted by wild-type BALB/c recipients treated with CTLA4Ig (MST > 100, n = 4). C, Pathological examination of the grafts (>100 days) showed severe chronic rejection and vasculopathy in PDL1-deficient hearts after CTLA4Ig treatment as compared with wild-type grafts which showed protection from chronic rejection and vasculopathy.

FIGURE 6

Effect of PDL1/L2 blockade on mixed allogeneic chimerism in a model of central tolerance. The percentage of H-2K^b chimeric donor cells in recipient peripheral blood 8 wk after chimerism induction was not affected regardless of treatments administered (controls: 47.23 ± 6.98, PDL1 blockade: 51.73 ± 7.8, and PDL2 blockade: 52.03 ± 2.46%). One representative example is shown in the top panel. Data are representative of four independent experiments and indicate the mean of duplicate results in each experiment.

FIGURE 7

Differential effect of PDL1/L2 blockade on allograft tolerance induced by mixed allogeneic chimerism based on a nonmyeloablative regimen using CD40-CD154 blockade versus tolerance induced by CD40-CD154 blockade alone. A, Fully MHC-mismatched skin allografts are accepted by conditioned transplant recipients. Tolerance is abrogated neither by PDL1 nor PDL2 blockade. B, Tolerance to fully MHC-mismatched cardiac allografts induced by CD40-CD154 blockade is only abrogated by PDL1, but not PDL2 blockade.